Statistics Questions Based Assignment

Question 1

The procedure commenced by checking the outliers of the data presented. Based on the findings generated from the SPSS software, it was identified that there was only single outlier in the overall model. The outlier is represented in the form of Boxplot. Asterisks or other symbols were displayed using the boxplots on the graph for indicating comprehensively when datasets comprise of outliers. It has been observed that these graphs utilize the interquartile method when links with other fences for finding outliers. From the data file, it is observed that the outlier was quite distinguish as compared to the typical data value. Based on this finding, it can be stated that the data were normally distributed and fulfilled all the assumptions.

Figure 1

Outliers

Following table present descriptive statistics undertaking minimum, maximum, mean value, and standard deviation. There were no missing data observed in the descriptive statistics. The mean value for the mental and physical training was 2.50 and standard deviation was 1.122. In addition, the mean and standard deviation values for perceived stress scale were 25.30 and 15.27.

Table 1

Descriptive Statistics

Figure 2

Line Numbers showing Mean Values

One-way ANOVA was used to examine the post-test scores based on perceived stress scale. The findings have shown that there is a positive and significant difference in the post-test scores based on the perceived stress scale (Table 2). Moreover, the findings have indicated that there was a significant difference between the scores of waitlist control group with respect to mental and physical training program. In addition, it was observed that there was a significant impact of waitlist control group and mindfulness training alone. Lastly, the findings have also indicated that mindfulness training alone and waitlist control group and mindfulness training plan; and waitlist control group and mental and physical program. The confidence interval for all the relationships showed exact mean values in the predefined range.

Table 2

One-Way ANOVA

Table 3

Post-Hoc Test (Bonferroni)

It has been observed from the findings that the satisfaction with the profession emerges to be associated with satisfaction with workload, organizational environment, and satisfaction with professional associations; all factors that contribute to overall subjective wellbeing. The most adequate approaches for enhancing mental and physical training, reducing stress, and enhancing mindfulness include maximizing those conditions in which there are improved chances to appreciate or identify the work completed by employee; improving those conditions in which there are encouraging supervisors; and elevated chances for professional development. Work flexibility and supporting an environment where coworkers are mutually encouraging are the areas that can be considered. It is beneficial for maximizing those conditions in which social workers assume that they are addressing with workload as well as handling workload with respect to helping them maintaining boundaries between their personal and professional lives, mitigating bureaucratic work contexts, and offering flexible time off from work. Professional associations an improve areas such as provision of assistance on ethical and legal issues, helping the enhancement of working conditions, and promoting the social work profession, which can bring significant effects on the individual’s lives. It has been observed that improving social wellbeing of an individual can further be improved by benefitting the social workers and the societies they deal with.

Question 2

Missing Values Identification

Table 4 shows descriptive statistics in order to identify missing values related to each variable. It has been identified from the column representing N that all the variables represented complete observations instead of post-test variable, which showed only 63 observations and 7 missing values.

Table 4

Descriptive Statistics

Chi-square statistics is performed for identifying the relationship between group and gender. The group is divided into patients receiving treatment and healthy patients as control group (Table 5). The findings have shown chi-square statistics for 70 values as 0.168, which was insignificant at 5% level of significance. Both variables showed complete observations and none of the variable revealed any missing value.

Table 5

Chi-Square Statistics based on Group and Gender

Best ways for handling Missing values

The effective solution is to reduce missing values when the data were being collected. It has been observed and suggested that a researcher should demonstrate how cases are reduced from test and the percentage of observations reduced by different tactics when working with missing observations. In addition, researchers should update information on why an individual is having a missing value. Differentiating what should and should not be imputed is often impossible with a single code for every missing value category.

Assigning or imputing a value may be accountable if a do not know response is interpretable underlie between agree and disagree on a continuum issue. Otherwise, it will become problematic. The new approaches for working with missing values were effective tools that can be misused if values were imputed by researchers for individuals who should be excluded from the analysis.

In addition, it has been observed that all potential mechanism variables should explain missingness even when these are not encompassed in the analysis phase (Meng, 2000; Rubin, 1996). All variables were included both outcomes and determinants in the model at the imputation stage. This association must be implemented in the imputation phase if the dependent variable is associated to an independent variable. The parameter estimates for an investigation variable that is not encompassed will be biased downward in the imputation step (King et al., 2001; Meng, 1995; Rubin, 1996). It is complicated for identifying whether full information maximum likelihood estimation or multiple imputation is effective, but both are major developments over conventional tactics. It has been observed that both work effectively on large and complicated samples.

In the dataset, the approaches are not primarily associated with the forecasting of values for particular individuals (Perkins et al., 2019). Some of the imputation methods emphasized in this assignment where a researcher with statistical knowledge can effectively integrate some of the imputation methods (Anselmi, Robusto & Cristante, 2018). In general, any approach that can unequally regulate spaced and distinguish numbers of observations on each individual can organize MCAR data. The accommodation of data that is MAR can be effectively handled through likelihood-based longitudinal methods. On the other hand, likelihood-based longitudinal methods are not effective for NMAR data.

Cnaan, Laird & Slasor (1997) have indicated that statistical software packages are competent with some of more complicated imputation method explained including the SAS procedure, BMDP procedure, MIXED, and the SPSS Missing Values Analysis procedure. both single imputation and multiple imputation were implemented based on specialist software such as SOLAS (Zhu, Wang & Samworth, 2019). This software can handle data in several different formats such as SPSS, SAS, and S-Plus datasets. It has been observed that the researcher has several options with missing values when choosing how to handle with this common issue (McNeish, 2017). It might be beneficial for performing a sensitivity analysis based on different tactics in order to handle the missing data for evaluating the robustness and appropriateness of the findings.